Andrea Passerini

GlanceNets: Interpretabile, Leak-proof Concept-based Models

2022 · arXiv (Cornell University)

There is growing interest in concept-based models (CBMs) that combine high-performance and interpretability by acquiring and reasoning with a vocabulary of high-level concepts. A key requirement is that the concepts be interpretable. Existing CBMs tackle …

Neuro-Symbolic Reasoning Shortcuts: Mitigation Strategies and their Limitations

2023 · arXiv (Cornell University)

Neuro-symbolic predictors learn a mapping from sub-symbolic inputs to higher-level concepts and then carry out (probabilistic) logical inference on this intermediate representation. This setup offers clear advantages in terms of consistency to symbolic prior knowledge, …

Learning to Guide Human Experts via Personalized Large Language Models

2023 · arXiv (Cornell University)

In learning to defer, a predictor identifies risky decisions and defers them to a human expert. One key issue with this setup is that the expert may end up over-relying on the machine's decisions, due …

You Don't Bring Me Flowers: Mitigating Unwanted Recommendations Through Conformal Risk Control

2025

Recommenders are significantly shaping online information consumption.While effective at personalizing content, these systems increasingly face criticism for propagating irrelevant, unwanted, and even harmful recommendations.Such content degrades user satisfaction and contributes to significant societal issues, including …

Legal Argument Mining: Recent Trends and Open Challenges

2025 · Archivio istituzionale della ricerca (Alma Mater Studiorum Università di Bologna)

This paper presents a brief survey of recent trends in legal argument mining, focusing on the early use of large language models in this subfield. As legal texts, especially judicial decisions, increase in volume and …

If Concept Bottleneck ARE THE QUESTION, ARE FOUNDATION MODELS THE ANSWER?

2025 · Zenodo (CERN European Organization for Nuclear Research)

Concept Bottleneck Models (CBMs) are neural networks designed to conjoin high performance withante-hoc interpretability. CBMs work by first mapping inputs (e.g., images) to high-level concepts(e.g., visible objects and their properties) and then use these to …